ut3_tucker_weights_sharing_residual#

t3toolbox.backend.sharing.ut3_tucker_weights_sharing_residual(weights_data, sharing)#
def ut3_tucker_weights_sharing_residual(
        weights_data:   typ.Tuple[
            NDArray,             # tucker_weight_supercore, shape=(d,)+stack+(n,)
            NDArray,             # tt_weight_supercore,     shape=(d+1,)+stack+(r,)
            typ.Tuple[NDArray, NDArray],  # (tucker_edge_mask, tt_edge_mask), HOST bool, static
        ],
        sharing:        typ.Sequence,   # len=d, static; one hashable group label per mode
) -> NDArray:  # shape = stack_shape; max relative MASKED Tucker-weight deviation per stack element

The uniform twin of t3_tucker_weights_sharing_residual(): the comparison runs on the MASKED Tucker weight vectors (padding is don’t-care), and unequal group rank masks raise (structural – unequal ranks cannot carry one shared weight).

Parameters:
  • weights_data (t3toolbox.backend.common.typ.Tuple[NDArray, NDArray, t3toolbox.backend.common.typ.Tuple[NDArray, NDArray]])

  • sharing (t3toolbox.backend.common.typ.Sequence)

Return type:

NDArray